Work

View-relation constrained global representation learning for multi-view-based 3D object recognition